Two ways to feed ScrapeBox
ScrapeBox can harvest proxies from a custom source or load them from a file. The harvester suits short sessions: you harvest the ProxyMix link, move the result to the proxy list and start. Long runs work better with a file that a scheduler rewrites every 10 minutes, so you reload it whenever the error rate grows.
For a package, put the address of your ScrapeBox computer on the allowed list in the ProxyMix panel before the first harvest.
Step by step
- Open the Proxy Harvester and add your list link as a custom source.
- Harvest, then move the result to the proxy list of the main window.
- Mark the proxies as SOCKS in the proxy settings, or use the lines as HTTP.
- Set the maximum connections no higher than the package threads.
- For runs longer than an hour, harvest again every 10 minutes or load the file that the scheduler keeps fresh.
A file that stays fresh
The crontab line below downloads the list every 10 minutes and replaces the file only after a complete download. On Windows, Task Scheduler runs the same download on the same schedule.
*/10 * * * * curl -fsS "YOUR_LIST_LINK" -o /opt/proxies/list.tmp && mv /opt/proxies/list.tmp /opt/proxies/proxies.txtConnections and search engines
ScrapeBox opens many connections per search engine, and engines block fast. Keep the connections per engine moderate, take a bigger list and let the harvester spread requests across more addresses. The package threads cap the total across all engines.
Skip the built-in proxy test after each harvest. The pool serves working addresses, and testing thousands of lines only burns threads.
Tips
- Many proxies marked as failed? Lower the connections first: extra connections get refused and look like dead lines.
- Use the CIS pool for Yandex and World MIX for Google.
- Save harvested lists with the date in the file name, so you never load a stale copy.
ScrapeBox jobs that suit the list
- Harvesting URLs from search results for a list of keywords.
- Checking which pages a search engine has indexed.
- Collecting keyword suggestions and checking backlinks.
- Mass checks of response codes for URL lists.
All of them send many short requests. The more lines in the list, the fewer requests each address takes.
Residential lines in ScrapeBox
ScrapeBox reads proxies with a login as host:port:login:password, the same form the panel gives, so paste the lines through the clipboard as they are. Keep them as HTTP proxies: create HTTP lines in the panel and leave the SOCKS mark off.
A rotating “Country” line spreads harvester requests over home addresses when a search engine blocks data center ranges. Harvesting downloads many result pages, and residential traffic is billed per gigabyte, so mass harvests stay cheaper on a package.
Test before buying
Once your account is verified, ask support for a free hour and run a real harvest. Thread limits are on the prices page, and the curl guide shows the same list in shell scripts.